Ovarian cancer is usually diagnosed at an advanced stage, at which point only about a quarter of patients will survive for at least five years. There are currently no screening tests to detect the disease earlier when survival rates can be much higher.
Now, a research team has utilized artificial intelligence to create a more accurate test to diagnose ovarian cancer earlier.
The investigators, hailing from Brigham and Women’s Hospital as well as the Dana-Farber Cancer Institute, conducted an experiment to identify a non-invasive tool that would be more sensitive and specific in detecting true cases of early-stage disease in the blood.
This test was split into two parts.
First, the scientists analyzed a series of molecules called microRNAs, which are small, non-coding pieces of genetic material. The team then sequenced microRNAs in blood samples from 135 women prior to surgery or chemotherapy that would lay a foundation for a computer program to look for microRNA differences between cases of ovarian cancer and cases of benign tumors, non-invasive tumors, and healthy tissue.
Adopting this machine learning approach enabled the researchers to leverage large amounts of microRNA data to yield different predictive models.
Next, this sequencing model was used in an independent group of 44 women to verify the accuracy of this new diagnostic, which was then followed by gauging the model’s sensitivity and specificity on 859 patient samples.
Results indicated the new technique was far better at predicting the presence of the disease versus an ultrasound test. Almost 100 percent of abnormal results using the microRNA variant represented ovarian cancer while the ultrasound approach found fewer than five percent of abnormal test results signified ovarian cancer.
“The key is that this test is very unlikely to misdiagnose ovarian cancer and give a positive signal when there is no malignant tumor. This is the hallmark of an effective diagnostic test,” said Dipanjan Chowdhury, Ph.D., the senior author of this study and head of the division of Radiation and Genomic Stability in the Department of Radiation Oncology at Dana Farber.
More work will need to be done in order to move this test out of the lab and into the clinic. The researchers will need to collect longitudinal samples following women over time so the program can learn how to verify how these microRNA signatures can change.
